In this paper we analyse the behaviour of departmental performance in English secondary schools over time. Out unique database enables us to estimate value added at departmental level and across a number of years. We are thus able to examine whether departmental effects are stable over time, an important issue and one where there is no clear theoretical prediction. A priori we might expect that a particularly effective department may be able to sustain a relatively high value-added effect over time. However, in an increasingly results-orientated environment, less effective departments will continually face pressures to raise their own relative performance. We investigate the size and stability of departmental effects over time during a period in which extensions of parental choice and annual publication of school performance tables had significantly increased competitive pressures in English secondary schooling (Johnson, 2004). Our sample is over 600 schools from the Curriculum, Evaluation and Management Centre's YELLIS database covering the period 1994-2002. This data provides us with information on student characteristics, including a measure of student ability, and student expected and actual attainment across a range of optional subjects in English # Corresponding author, Centre for the Economics of Education, London School of Economics and Political Science, Houghton Street, London, WC2A, Tel: 02078523578, e-mail: s.telhaj@lse.ac.uk) † Curriculum, Evaluation and Management Centre, University of Durham * Institute for Education Policy Research, Staffordshire University Business School secondary schools. Through the use of multi-level modelling we are able, after having adjusted for student characteristics, to distinguish between school and departmental influences on value-added. Our data enables us to examine the behaviour of these influences over time and across a range of different types of schools. Our empirical is enables us to examine the argument of Adnett and Davies (2005) that policies encouraging greater competition within schools may be more effective than the current emphasis on increased competition with schools. This argument challenges one of the key propositions underlying the school choice critique of public schooling. Our empirical results therefore have the potential to make a significant contribution to current policy debates in Europe and elsewhere.